37 replies to this topic

[justcallme A S H]

Waste of Ammo, on 05 October 2017 - 03:27 AM, said:

can Sync drops of whole groups (abuse of MM) in quick play be the root cause of the stomp-games ??

No.

Because it's so rare of a thing and people end up on both sides usually anyhow.

[Ghogiel]

MM is better... but the pop is low af right now so it's making as many bad matches as ever for most of the day.

[Jingseng]

this looks like a job for..... iterative machine learning =p It would be ideally how you account for the buckets (ha) of variables.

It's also well beyond pgi's budget and means.

[PyckenZot]

Let me tell you a story from another game.

WoWS has a new campaign to win some sort of Russian battleship. First mission in this series was to do X amount of damage, win the drop and survive. Statistically speaking, should've been feasible in 3-5 drops,...

It took me 3,5 HOURS! For some godforsaken reason, I always reached my damage score usually even survived. But every single drop the team was able to snatch defeat from the jaws of victory.

As a cherry on top, the first game I won after all that playing, I didn't reach my damage goal.

TL : DR
RNGesus can be a ******* ****!

Edited by PyckenZot, 05 October 2017 - 05:34 AM.

[SFC174]

Tarl Cabot, on 05 October 2017 - 03:19 AM, said:

Remember PGI may have tighten up the MM parameters but it did not change the current PSR parameters that covers a long period of time without any sort of reset/realignment.

But really, how is this really different from Competitive play and their lopsided wins there, and not just this year but last year? Actual teams where both sides are not bringing joke builds, most are CLAN mechs, communication and some sort of gameplan on both sides, one less lance on each side, etc?

Is there a matchmaker for competitive teams? Not in the sense that we're talking about here. Competitive is about putting your best team together and fighting someone else's best, and really much of the match is already decided by who builds/trains the better team. The hope for a MM in quickplay is that it would create the most evenly matched teams possible for greatest enjoyment of the player base.

If you stomp someone in comp queue you did your job in building the better team and a stomp is a sign of your success (opponent's failure). Stomps in QP indicate the MM didn't do a great job of maximizing entertainment value (IMO).

[Big MO]

Agreed, there are so many variables an optimised MM would need to take it would be very difficult. Myself, for example, I am 50 years old with not the greatest eyesight. Drop with me in Viridian Bog at night, in the rain, and you are basically a mech down.

[The Mysterious Fox]

MWO is just built in a way that snowballs very easily, reversals are possible but usually once the tide comes, theres no stopping it

[Lykaon]

Most of the time when I see a a complete disaster of a match I can pin down the specific point in the game where my team messed up.

It's either failing to properly support a push resulting in casualties (usually heavy and assault mechs)

Failing to adapt tactics to counter enemy strengths (sitting still to long on a ranged map when the enemy mechs have the build advantage)

Losing right out of the gate (watching a team scatter to the four corners of a map on a skirmish)


Quick play has it's tropes... Players will do stupid things simply because it's what you do in quick play solos.

One of my all time favorites is River City and how the southern deployed team will run to G6.

G6 is a terrible possition to take. There is no cover from either flank or the rear no means of advancing straight forward and the left flank is difficult to manuver assaults through to get out of the water.

This leaves a right flank push as an obvious tactic to take but you will be faced with enemy with both cover and high ground as well as the capacity to fall back into a second defensive line if somehow you do succeed with the initial push.

And then there is Tourmaline and how players just seem to love running into the basin of death and triggering a choke point scenario for any of their future forward advances.

Choosing Frozen city domination and failing to actually take the objective because of how it's in an intimidatingly low basin.

Choosing HPG manifold on most modes and not pushing to take the top.

Choosing Veridian Bog and playing that stupid "go-cart bumpercar" version of NASCAR on the central mesa were one team is literally tens of meters away on the other side of the mesa poke left then right then left left again back to right....

[CFC Conky]

Full disclosure, I'm a Tier 5 player and not looking for any unfair advantage in the game.

At his time, stomps are a fact of life in MWO, making the game pretty brutal on new players which in turn may be a factor in the low player population.

At the risk of being forum-stomped, here's a crazy idea: Have the MM include all Tiers in QP, but the higher your Tier, the less damage your weapons do, say, just for sh*ts and giggles, 100% for Tier 5, 80 for T4, 70 for T3, 60 for T2 and 50 for T1.

High Tier players would still be very dangerous but new players might be able to survive long enough to learn something and begin enjoying the game. Sure, T1's could just start an alt account and clean up as a T5 but we're probably seeing that already.

The idea is to attract and retain new players to MWO.

Flame away,
CFC Conky

Edited by CFC Conky, 05 October 2017 - 04:34 PM.

[justcallme A S H]

Neutering good players with damage/hp reduction is the way games die. Good players spend time, getting good.

No thanks.

[Wil McCullough]

CFC Conky, on 05 October 2017 - 04:33 PM, said:

Full disclosure, I'm a Tier 5 player and not looking for any unfair advantage in the game.

At his time, stomps are a fact of life in MWO, making the game pretty brutal on new players which in turn may be a factor in the low player population.

At the risk of being forum-stomped, here's a crazy idea: Have the MM include all Tiers in QP, but the higher your Tier, the less damage your weapons do, say, just for sh*ts and giggles, 100% for Tier 5, 80 for T4, 70 for T3, 60 for T2 and 50 for T1.

High Tier players would still be very dangerous but new players might be able to survive long enough to learn something and begin enjoying the game. Sure, T1's could just start an alt account and clean up as a T5 but we're probably seeing that already.

The idea is to attract and retain new players to MWO.

Flame away,
CFC Conky

Imagine the tears on the forums. "Screw hitreg! I put a full burn into his ct and it only turned yellow! Wehhhhhhh!"

Also, i can already foresee the complaints of t1 pilots making alts to stomp potatoes. Nothing beats having double your damage in t5 when you already play like a t1. The salt and abysmal npe will drive this game into the ground. Glorious.

Edited by Wil McCullough, 05 October 2017 - 05:54 PM.

[Tarl Cabot]

SFC174, on 05 October 2017 - 05:34 AM, said:

Is there a matchmaker for competitive teams? Not in the sense that we're talking about here. Competitive is about putting your best team together and fighting someone else's best, and really much of the match is already decided by who builds/trains the better team. The hope for a MM in quickplay is that it would create the most evenly matched teams possible for greatest enjoyment of the player base.

If you stomp someone in comp queue you did your job in building the better team and a stomp is a sign of your success (opponent's failure). Stomps in QP indicate the MM didn't do a great job of maximizing entertainment value (IMO).

It is all of it is about team work. What type of algorithm could PSR/MM use for that? There isn't one. At least in Competition Play the players have banded together and actually working together. Group queue is a mix, one extreme the smaller the groups the closer that side is to the Solo queue while the side with the larger group is closer to competitive.

MM/PSR can not fix stupidity, it is all about me, I will play it the way I see fit, etc, etc.

[CFC Conky]

Thanks for the responses. Fair enough, with a couple of caveats.

First, while nerfing good players can kill a game, so can a lack of new players, but I get your point ASH.

Second, a T1 player using an alt account to play in T5 and stomp new players is already happening and if one wanted to stay in T5, well, that's a lot of alt accounts to create and a lot of trouble to go through just to pad your stats.

To express my original point in a different way, I don't think the 'suck it up buttercup, just git gud' approach to the npe is good for the longevity of the game, even if it's been this way from the beginning. I also said it should only be for QP, the harder modes of the game should be just that, harder.

To borrow an example from flight sims, specifically the IL2 Bo(X) series, navigation, situational awareness and keeping an opponent in sight is a big challenge for new players so some of the servers have player icons that are always visible. Experienced players view this as easy mode but it can ease a new player into the game; they can concentrate more on handling their aircraft, then learn to fight them, which are two very different things. Ideally, once experience is gained, the player can move on to no-icon servers for the full experience of the sim.

Good hunting,
CFC Conky

[SFC174]

Tarl Cabot, on 05 October 2017 - 06:06 PM, said:

It is all of it is about team work. What type of algorithm could PSR/MM use for that? There isn't one. At least in Competition Play the players have banded together and actually working together. Group queue is a mix, one extreme the smaller the groups the closer that side is to the Solo queue while the side with the larger group is closer to competitive.

MM/PSR can not fix stupidity, it is all about me, I will play it the way I see fit, etc, etc.

A proper player ranking system would take into account teamwork. Players who do a good job leading teams (having good communication skills, map awareness and tactical decision making) would generally have a higher win rate, and probably higher match scores as well. Players who follow direction well will also tend to average higher. OTOH, players who do their own thing will hurt their team more often and on average have lower WLR and match scores.

So, IMO, a good ranking/MM system will clearly reward teamwork, and as you get to higher tiers you'd expect better and better cooperation, even in solo quickplay.

[Mainhunter]

A pure player ranking by k/d stats alone wouldn't do it. You should take also into account the value of the just used Mech and how well the player did with it in the past.

[MischiefSC]

CFC Conky, on 05 October 2017 - 06:24 PM, said:

Thanks for the responses. Fair enough, with a couple of caveats.

First, while nerfing good players can kill a game, so can a lack of new players, but I get your point ASH.

Second, a T1 player using an alt account to play in T5 and stomp new players is already happening and if one wanted to stay in T5, well, that's a lot of alt accounts to create and a lot of trouble to go through just to pad your stats.

To express my original point in a different way, I don't think the 'suck it up buttercup, just git gud' approach to the npe is good for the longevity of the game, even if it's been this way from the beginning. I also said it should only be for QP, the harder modes of the game should be just that, harder.

To borrow an example from flight sims, specifically the IL2 Bo(X) series, navigation, situational awareness and keeping an opponent in sight is a big challenge for new players so some of the servers have player icons that are always visible. Experienced players view this as easy mode but it can ease a new player into the game; they can concentrate more on handling their aircraft, then learn to fight them, which are two very different things. Ideally, once experience is gained, the player can move on to no-icon servers for the full experience of the sim.

Good hunting,
CFC Conky

You don't nerf skill, but you absolutely do want the QP environment as skill-segregated as you can get it. The problem there however is that you get new players who never learn good habits because they're only playing with/against bads.

Ideally you want a bit of a mix; teams all consisting of 4 T1/T2 players and the rest T5-3. Even better is when the experienced players call the drop or at least help with communication.

You don't want everyone who comes to the game to have to reinvent the wheel, the combustion engine and automotive mechanics to learn to drive. You want them to learn from someone who already knows.

With the current tier system though that's hard because tier /= skill. An Elo system that balanced both player, mech and loadout would be the best and also help give people feedback on the realities of what works best. This game has a lot of things to wrap your head around to start to GIT GUD and there's almost no 'official' sources for most of them.